Accueil | Sciences Po Centre d'études européennes

> Renaud Dehousse

The Politics of Delegation in

the European Union

LES CAHIERS EUROPEENS DE SCIENCES PO.

> N° 4/2013

Renaud Dehousse – The Politics of Delegation in the European Union

Les Cahiers européens de Sciences Po. – n° 04/2013

2

Les Cahiers européens de Sciences Po. n° 04/2013 Décembre 2013

RENAUD DEHOUSSE

The Politics of Delegation in the European Union

Renaud Dehousse holds a Jean Monnet Chair in European Union Law and Politics at Sciences Po, Paris, where he directs the Centre d'études européennes. He has published extensively on the institutions and political system of the European Union, and on EU public policy. He is advisor at Notre Europe and has participated in a number of high-level groups on the reform of European institutions set up by the European Commission and the French government.

Contact: [email protected]

Citation : Renaud Dehousse (2013), “The Politics of Delegation in the European Union”, Les Cahiers européens de Sciences Po, n° 4.



3


Keywords: Agency theory, Institutions, Economic integration, Integration theory, European Commission

Abstract:

Delegation of powers to supranational institutions, once a hallmark of European integration, is increasingly contested by national leaders as well as by the public opinion. At the same time, recent developments suggest that in turbulent times, the technique remains widely used. This article purports to explain this apparent paradox. It proposes a reading of the principal-agent principles that takes into account specificities of the EU system, such as the absence of a strong centre of power or the significant degree of mistrust that may exist among national governments. It argues that the competition between multiple principals may ultimately result in different models of delegation. Résumé: La délégation des pouvoirs à des institutions supranationales, qui était une caractéristique principale de l’intégration européenne, est de plus en plus contestée par les leaders nationaux ainsi que par l’opinion publique. Cependant, les évolutions récentes tendent à montrer que dans les périodes de crise actuelles, cette technique est encore largement utilisée. Cet article vise à expliquer ce paradoxe apparent. Il propose une lecture des principes du principal-agent qui prennent en compte les spécificités du système de l’UE, telles que l’absence d’un lieu de pouvoir fort ou la grande méfiance qui règne entre les gouvernements nationaux. Il soutient que la concurrence entre un grand nombre de « principaux » peut donner lieu à terme à différents modèles de délégation.



4


When Robert Schuman presented his blueprint for what was to become the Coal and Steel Community (ECSC) on the 9th of May 1950, it was clear to most observers that the proposal was a radical departure from the classical architecture of international organizations. For the new organization to effectively address the problems of the time, Schuman explained, it was to enjoy unprecedented powers. To make ‘any war between France and Germany (…) not merely unthinkable, but materially impossible’, as was the ambitious objective of the proposed initiative, major innovations were needed: the pooling of coal and steel production required a complete transfer of authority to a new body, the High Authority, which was to act independently from national governments and make decisions that would be directly enforceable by member countries.

Delegation of powers to supranational institutions – a concept used in the Schuman

Declaration itself – would be the hallmark of the new system. Political actors often refer to the ‘Community method’, developed on the basis of Schuman’s blueprint, as the operating system of the European Union (EU). Textbooks on EU institutions and politics routinely start by listing the features that distinguish the EU from other international organizations, while comparisons between the European Union and other models of regional integration stress the uniqueness of the former1. In the press, the European Commission, successor to the High Authority, is often described as a Leviathan, whose primary concern, like any bureaucracy, is to maximize its power. At the same time, however, the Community model has recently come under attack. Since the Maastricht Treaty, support for the integration process has been waning, and governments have proven to be increasingly reluctant to accept further transfers of powers to European institutions. The sovereign debt crisis that has shaken the Eurozone in recent years has spawned a revisionist approach. National leaders have become more assertive. In a much-noticed address delivered in 2010 at the College of Europe in Bruges, Chancellor Merkel heralded the emergence of a new “Union Method”. Despite this being a hazy notion, the political message was clear: new modes of governance, more adapted to the needs of our time, had to be considered. As was often the case, French President Nicolas Sarkozy was more blunt:

‘The crisis has prompted the heads of state and government to assume greater responsibilities because at the end of the day, they alone have the democratic legitimacy to make decisions. European integration will be intergovernmental because Europe will have to make strategic, political choices.’2

It will be obvious to most readers that there are huge differences between the context

in which the supranational model was first invented and that in which the European Union

1 R. Severino, ASEAN , Singapore: Institute of South East Asia Studies, 2008. 2 ‘La crise a poussé les chefs d'État et de gouvernement à assumer des responsabilités croissantes parce qu'au fond, eux seuls disposaient de la légitimité démocratique qui leur permettait de décider. C’est par l'intergouvernemental que passera l'intégration européenne parce que l'Europe va devoir faire des choix stratégiques, des choix politiques.’ Speech delivered in Toulon on November 1st, 2011. Author’s translation.



5

currently operates. Initially envisaged as a Western citadel, the EU has expanded and now covers most of the continent. The Cold War is over, and fear of foreign invasions has given way to fear of terrorist threats. Although some European countries are going through the most severe economic crisis since the 1920s, their current situation has little in common with the dramatic conditions of the post-World War II period. At the same time, they are confronted with new problems, such as competition from emerging countries, immigration, and climate change. These issues could have a dramatic impact on European economies and societies. New political elites, mostly born after 1945, do not seem to view integration in the same benevolent light as their predecessors.

This contribution aims to shed light on a simple question: has the balance of power

between the European Union and its member states really changed in response to these transformations, or are we simply witnessing a rhetorical change? The past can serve as a guide to answering this question. We must first understand the reasons that prompted the drafters of the first treaties to propose unprecedented transfers of powers, and the signatory countries to accept these transfers. Several distinct but clearly related facets of the problem need to be analyzed: the types of motivation that led governments to accept such transfers of powers, the authority enjoyed by supranational agents, and how these agents can be organized to secure the governments’ trust. To this end, we will largely rely on the principal-agent (P-A) model, which is widely used to discuss delegations of power, and we will try to confirm its predictions with historical evidence. We will also address what has been described as a major failing in the literature on the P-A model in the EU, namely the insufficient attention3 that has been given to a variety of issues arising from the existence of multiple principals. In a second stage, we will look at how the very same delegation issue might arise in the European Union of the 21st century. This chapter is therefore organized as follows. Part 1 presents the basic elements of the principal-agent model. The structural concerns that have shaped the politics of delegation in the European Union, notably the polycentric character of the EU system, will be analyzed in part 2. We will then discuss the two patterns of delegation that have emerged: the ‘full delegation’ model, in which EU institutions enjoy substantial discretion (Part 3), and the weaker variation that developed from the 1990s onward (Part 4). Part 5 discusses the changes that took place after the Maastricht treaty – widely regarded as a turning point in the history of European institutions – and in response to the sovereign debt crisis, with a view to assessing the importance of delegation in today’s European Union.

1. A Snapshot of the Principal-Agent Model

When discussing the delegation problem, studies of institutional design often rely on a rational choice model first developed to analyze relationships between the US Congress and administrative agencies, and then transposed to the study of international regimes and

3 H. Kassim and A. Menon, ‘The Principal-Agent Approach and the Study of the European Union: Promise Unfulfilled?’, Journal of European Public Policy, 10(1), 2003, p. 121-139 at 134.



6

ultimately of the European Union.4 In this model, one or more principals expressly or implicitly transfer part of their authority to an agent, who will work on their behalf, and whose work will be controlled. The model spells out a series of reasons that may justify resorting to such a technique; it also envisages a series of mechanisms that can be used to induce the agent to act in the principals’ interest. It can therefore be of great help in analyzing the politics of delegation, be it at the time the principle of a delegation of authority is agreed upon, or at the time of implementation. Why Delegate?

The Principal-Agent model is derived from transaction cost economics, which identify a variety of costs – called transaction costs – that may be incurred in economic exchanges. Some of these costs are informational (availability and price of commodities); others are linked to the transaction itself (bargaining costs); and yet others are linked to the transaction’s implementation: will the other party do what has been agreed? And what if they don’t? (enforcement costs).

As mentioned above, these notions have been transposed to the study of public policy,

to analyze relationships between the executive and the legislature, or between different actors. It has been emphasized, for instance, that policy-making could entail significant information costs. The issues that are addressed are often complex: what is the impact of industrial activity on climate change? Is a drug safe and efficient? The crafting of efficient responses to such questions may require a technical expertise that lawmakers do not necessarily possess. They may therefore find it convenient to designate specialized institutions staffed with experts to provide them with policy-relevant information, or even to make decisions on certain issues. Bargaining costs may exist in relation to any decision. It may therefore be costly to fully detail what has to be done and how. Legislative decisions will very often establish a number of basic principles and procedures governing the implementation of these principles: who will enact secondary legislation, how potential interpretation problems will be resolved, and what will happen in the event of an unanticipated problem. If the initial agreement is vague and future decisions can be expected to be a source of disputes, parties may choose to entrust an agent – be it a judicial or an executive body – to fill in the details and/or adjudicate potential disputes.5

Another recurrent theme in the principal-agent literature is the credibility of policy

commitments. Policymakers who are keen to adopt a given measure cannot be sure that a new majority with a different policy agenda will not subsequently reverse the measure. The resulting lack of long-term perspective may be problematic in areas where huge investments are required. Telecommunications are a good case in point: the industry will be reluctant to invest the large amounts necessary to upgrade to the latest technology unless it has

4 See e.g. S. Hix, The Political System of the European Union, 2nd. ed., Basingstoke: Palgrave Macmillan, 2005; M. Pollack, The Engines of European Integration, Delegation, Agency and Agenda-Setting in the EU, Oxford University Press, 2003. 5 On the notion of incomplete contracting, see O. Williamson, The Economic Institutions of Capitalism, New York: The Free Press, 1985.



7

sufficient guarantees that a given policy will be maintained long enough for its investment to pay off.6 Similar credibility problems may occur as politicians often find it difficult to stick to “virtuous” commitments they have undertaken: fiscal austerity or anti-inflation policies may be good for the economy in the long run, but the governments that adopt them tend to be penalized by voters. In both cases, a regulatory agency or a central bank protected from political pressure are likely to enjoy greater credibility than politicians who may be influenced by electoral concerns.7 Endowing such non-majoritarian bodies with clear objectives and discretionary powers to pursue them has often appeared as a way to guarantee the continuity of a given policy.

While the above cases mainly refer to situations in which a single principal might find it convenient to resort to delegation, a similar logic may apply in a situation involving several parties, of which each has its own interests and preferences. In this multiple-principals variant, delegation may appear as a protection against the risk of defection of the other party. The problem is well known in international relations: why should a state fulfil its obligations if it has no guarantee that the other parties to an agreement will do likewise? To reduce uncertainty, monitoring or enforcement powers can be entrusted to neutral umpires, be they independent commissions, such as the United Nations Human Rights Committee, or judicial bodies proper, such as the European Court of Justice.

How much Discretion ?

As can be imagined, even when principals find delegation convenient for the above-

mentioned reasons, they often have a fairly precise idea of the kind of outcomes they find desirable, as well as of those they want to avoid. On the other hand, agents may have interests and preferences that differ from those of their principals, either because they do not share the same policy preferences, or because they develop interests of their own, such as the desire to see their own role recognized or preserved. They may be ‘captured’ by the interests they are supposed to regulate or by other political actors. Finally, principals may lack precise information about their agent’s activities.

Several types of instruments can be used to avoid such agency costs, i.e. the costs

that can be incurred when the agents’ behavior deviates from the principals’ expectations. A careful selection of agents is of course a simple way to avoid unwanted drifts: the appointment procedure is therefore likely to receive much attention, to ensure that the agency head has the right beliefs, be they pro-environment, pro-competition or pro-consumer. Another technique is to limit agents’ discretion: their powers can be limited to information-gathering, with principals retaining the right to decide, or their mandate can be well-defined, describing the kind of objectives they ought to pursue in precise terms. Various types of control mechanisms can be implemented, be it before (ex ante) or after (ex post) the

6 M. Thatcher, The Politics of Telecommunications: National Institutions, Convergence and Change in Britain and France, Oxford University Press, 1999. 7 G. Majone, Regulating Europe, London: Routledge, 1996.



8

agent takes action, to ensure that the agent will not deviate too widely from the aims of its mandate8; incentives can also be established to encourage compliance and deter deviations.

As can be seen, this approach analyses delegation problems in a purely rational way. It

emphasizes the functional motivations that may underpin the decision to delegate or the way delegation is engineered. Actors are supposed to behave as ‘rational egoists’ with a clear vision of their interests and of the most efficient way to achieve them. A historicized account or a more sociological approach highlighting the role of individual actors within each institution, would undoubtedly shed a different light, and in all likelihood provide a more fine-grained image of the same phenomena.9 Without going that far, we will try to illustrate the relevance of the model by discussing some examples borrowed from the institutional history of European integration. Before we proceed to show this, however, a few general remarks are in order.

2. A Political System without a Principal

Basic Principles

The analysis of European public policy often makes use of notions, concepts and theories developed in other settings. That is perfectly sensible, as comparison is one of the most useful methods in the social sciences. However, in doing so, one must not forget one of the key precepts of any comparative endeavour: the differences that may exist between the two situations that are being compared need to be duly taken into consideration. Many of the analytical tools used to make sense of EU policy-making have been fleshed out in the context of political systems in which there is no functional equivalent to the structural tension between the Union and its member states, or even between the national governments, that underpins much of EU politics. In other words, the European Union is not a nation-state, and that basic difference often has a strong influence on the way it operates.10 The remark might seem trivial but it is not, as this point is often neglected.

In relation to the principal-agent theory, for instance, I would argue that insufficient

attention has been paid to the key feature of the EU’s institutional structure, i.e. its character as a political system without a real principal. As is known, the construction of Europe is the product of joint decision by a number of countries that decided to pool their sovereignty in selected areas. Although European institutions have been endowed with powers that go beyond those of ‘ordinary’ international organizations in several respects, the member states of the EU remain the ‘masters of the treaties’: the latter can only be modified by unanimous decisions ratified in each country. This is not a mere formality, as was clearly demonstrated

8 M. Bovens, D.Curtin and Paul ‘t Hart (eds.), The Ral World of EU Accountability: What Deficit ?,Oxford University Press, 2010. 9 J. Lagroye and M. Offerlé (ed.), Sociologie de l’institution, Paris: Belin, 2011. 10 See R. Dehousse, Comparing National Law and EC Law: The Problem of the Level of Analysis, American Journal of Comparative Law, vol. 47, 1994, p. 201-221, for a more developed analysis.



9

by the lengthy and cumbersome revision process that ended with the ratification of the Lisbon Treaty: almost ten years elapsed between the launch of the reform process and the coming into force of the new treaty.

The institutional architecture of the EU, and its character as an international

organization, are not accidental: they reflect one of the cornerstones of the post-World War II architecture of Europe. European integration is indeed an unprecedented attempt to establish a new form of continental order without recreating the hierarchical power structure that characterizes most nation-states. Balance and consensus are the main features of this new system. Balance among the countries that decided to join the Community, and later the Union, has always been a central concern: no country is to be in a position to impose its views on the others. This anti-hegemonic ethos, together with the reliance on integration through law, is of course a major departure from earlier attempts to unite Europe by military force that were made by countries with imperial aspirations, such as France under Napoleon, or Nazi Germany. This anti-hegemonic character has shaped the institutional structure of the Union. Although important discretionary powers have been granted to an independent executive and an independent judiciary to ensure the effectiveness of the system, national interests cannot be completely ignored since the members of those bodies are appointed by national governments – even if they are vetted in various ways to ensure some consideration for the common interest. Balance among the institutions is another way to further this goal, since they each represent different kinds of interests – hence the attention the principle of ‘institutional balance’ has received in the case-law of the Court of Justice. 11

This broader political framework has fundamental implications for the way delegation

issues are handled in the EU system. First, delegation politics are shaped by the absence of a single principal. Second, the various principals do not share the same interests and preferences. These differences arguably constitute one of the main reasons why the principals are interested in delegation techniques.

Some examples

Consider for instance the situation at the time of the Schuman plan. Besides the

economic benefits expected from the proposed arrangement, one of the primary reasons why France supported the bold delegation scheme envisaged in the Schuman Declaration was ‘the fear that Germany, once freed from Allied control, would pose a renewed threat to French security owing to German industrial and consequently military superiority’.12 For French negotiators, one of the main objectives was to lock Germany into a cooperation scheme in which decisions would be both facilitated and more easily enforceable than they would in unanimity-based agreements, thanks to the part played by the supranational High

11 G. Majone, The European Community as a Regulatory State, Collected Courses of the Academy of European Law, Vol. V, Book I, The Hague: Kluwer, 1996. 12 B. Rittberger, Which Institutions for post-war Europe? Explaining the Institutional Design of Europe’s First Community, Journal of European Public Policy, 8(5), 2001, p. 673-708 at 686.



10

Authority. If European control of the coal and steel industry could be achieved, Allied control of the Ruhr and the Saar would lose much of its interest. For Germany too, the main priorities were of a political nature: at the time it was still an occupied country, and its major political and economic decisions had to be approved by Allied powers. Participation in the Coal and Steel Community was therefore perceived as a way to regain a greater share of sovereignty, which led Adenauer to insist inter alia that the Ruhr Authority be removed, and the Occupation Statute, weakened.13 For both countries, the establishment of an independent authority endowed with broad powers over strategically important sectors such as coal and steel, was probably a second-best choice. France would have preferred the status quo, but feared that time was running out, as the United States was pushing for Germany’s reinsertion in Europe to respond to the Soviet threat.14 For its part, Germany would probably have preferred to avoid the many constraints that came with membership in the ECSC, not least in terms of industrial concentration, but eventually accepted that this was the price to pay in order to regain a greater control of its destiny. In other words, for both of the key partners, delegation of powers to a neutral third party was an acceptable compromise, rather than an objective in its own right. But this was true only as long as the newly established body maintained the neutrality that played a key role in that compromise, as each of the would-be ‘principals’ knew that the other’s interests did not match its own.

A similar situation arose during negotiations on monetary union. Although the reasons

behind French support for the idea were manifold – a desire to consolidate the country’s commitment to low inflation, ideological support for European integration, as well as a geopolitical concern about binding a unified Germany to Europe – it is clear that one of the attractions of the single currency was that it could restore a degree of balance between France and Germany that had been disturbed by the Bundesbank’s domination in the European Monetary System of the 1970s and 1980s. French central bank governor Jacques de Larosière neatly summed this up in a 1990 statement:

‘Today I am the governor of a central bank who has decided, along with his nation, to follow fully the German monetary policy without having a vote on it. At least as part of a European central bank, I’ll have a vote.’15

French leadership also supported the idea of endowing the EU with some macro-

economic powers (“gouvernement économique”), with the clear ambition to achieve greater flexibility than the Bundesbank allowed. Meanwhile, Germany was radically opposed to anything that could weaken the Union’s commitment to monetary stability. It strongly insisted on the independence of the European Central Bank and on a lengthy transition based on economic criteria.

13 A. Milward, The Reconstruction of Western Europe1945-1951, London: Routledge, 1984 at 413. 14 A. Milward, op. cit., 13 at 378. 15 Washington Post, 25 October 1990; cited in A.Moravcsick, The Choice for Europe. Social Purpose and State Power from Messina to Maastricht, Ithaca: Cornell University Press, 1998, at 414.



11

As is known, the compromise eventually reached in Maastricht largely mirrored German preferences: the independence of the ECB is even better protected than that of the Bundesbank, since it has been given ‘constitutional’ status by its insertion in the Treaty; price stability was recognized as its main priority; bailouts were prohibited and strict convergence criteria were imposed. In exchange, however, Germany agreed to retain the old “one-state, one vote” rule in the ECB Governing Council – a major concession for a country that had for some time been dictating monetary policy to the rest of Western Europe. Subsequent events confirmed the importance of the latter point. When on 6 September 2012, the ECB President Mario Draghi made public the main components of its Outright Monetary Transactions (OMT) programme, aimed at preserving the common currency through purchases of sovereign bonds on the secondary market, he was forced to admit that the programme had met with opposition, with the German member of the Governing Council having voted against it. The fact that a decision of this importance, which some considered to run afoul the ‘no bail-out’ clause enshrined in the Treaty, was taken despite opposition from the hegemon that Germany had become (at least in macroeconomic policy) is a testament to the independence of the central bank.

Once again, therefore, the transfer of significant powers to an independent agent –

here the ECB – appears to have resulted from a compromise between different, if not conflicting, national preferences. No wonder, then, that assessments of the ECB’s behaviour widely varied from one capital to the next!

These two episodes underscore two key features of delegation politics in the EU

framework. First, since the European Union has a highly decentralized structure and is bereft of a clear centre of power, it entails dealing with different principals, each with their own views. Mistrust between the principals is a frequent motivation for delegating independent powers to a third party. Second, a key aspect of the relationship that develops between principals and their agents is linked to the relationship that exists between the principals. Because national governments are fully aware of the divergence between their respective preferences and are not naturally inclined to trust one another, they will tend to fear that other parties will defect or that the agent will be ‘captured’ by some other principal – a political drift in the language of agency theory. How they react to this fear will, however, vary according to several elements; this has implications for the kind of delegation that is used in the EU framework, as we will now explore.

The Two Logics of Delegation

In a seminal study published in 2001, Giandomenico Majone showed that it was necessary to distinguish between different kinds of delegation. The classical model discussed in the principal-agent literature emphasizes the willingness to reduce costs and improve the quality of decision-making as the main reasons for delegating power. The situation, Majone argues, is fundamentally different in cases where credibility problems are the main rationale for the delegation. When the credibility of member states’ commitment



12

needs to be enhanced, a more radical variant of the delegation principle will often be used. In this scheme, which he calls the ‘fiduciary principle’, responsibility in a given area is completely transferred to the delegate: a ‘complete transfer of political property rights takes place’.16 This situation can be compared to the common law institution of the trust, under which property is transferred to a person who is supposed to manage it for the benefit of another person. While both agency and trust rely on relations of trust and confidence, the two concepts are distinct. The trustee does not have a personal obligation to a principal, but must preserve a higher interest, which is why his or her independence is guaranteed.

This basic distinction between two kinds of rationales is at play here, for as Majone

perceptively noted, some of the problems that have arisen in the context of European integration do not fit well into the principal-agent model. In his own words: “the fiduciary principle has always been more important than agency as a mechanism for structuring the relations between the actors of European integration”.17 Indeed, I would argue that the main reason why mechanisms of full delegation to independent bodies have often been pursued in the European context is that there are many principals who are not particularly inclined to trust one another, so that delegation is seen as a way to ensure the effectiveness of their cooperation.

3. Full delegation in the EU

Consider for instance the organization of the European Central Bank created by the Maastricht Treaty. As mentioned earlier, for Germany to accept monetary union despite the fears voiced by the Bundesbank and the lack of enthusiasm in substantial portions of the public for jettisoning the Deutsche Mark, the independence of the new body had to be completely airtight. The ECB was therefore given a directorate that is often described as the “most federal” among European institutions, since it comprises only six members – far fewer than there are countries in the Eurozone. True, these people are appointed by the European Council acting by a qualified majority (Article 283 TFEU), and the national governments are not afraid of engaging in lengthy arms-twisting to secure a seat for one of their nationals. In 1998 this even resulted in an inglorious compromise, with France accepting the appointment of Wim Duisenberg as President only after he (unofficially) committed to step down after a few years to be replaced by a Frenchman (in this case Jean-Claude Trichet). This episode was widely seen as illustrating states’ grip on the appointment process, but it also shows the limits of state control over “political appointees”. Trichet, a well-known fiscal hawk who had pursued a tight monetary policy while he was governor of the French national bank, was publicly criticized for his conservatism by Jacques Chirac, the very same French president

16 G. Majone, The Two Logics of Delegation: Agency and Fiduciary Relations in EU Governance, European Union Politics, 2(1), 2001, p. 103-122, at. 106-107. 17 G. Majone, op. cit., at. 119.



13

who insisted on his appointment. Had France insisted on another candidate, it is doubtful that a deal could have been struck18.

Moreover, once they are in office, the independence of members of the directorate is

guaranteed by the strong language of Article 130, which is meant to shield them from national governments’ interference19. To protect the independence of the ECB, the same autonomy was extended to all Eurozone central bank governors, who together form the Bank’s Governing Council. Last but not least, these guarantees have been set in stone: they are part of the Treaty (like the statute of the European Central Bank, contained in a separate protocol) and can therefore not be changed without the consent of all Member countries –including, of course, Germany. In other words, the independence of the ECB is strongly protected, and this body has been given the power to conduct the monetary policy of the EU (Art. 282 §3).

Credibility concerns have played a central role in structuring the EU’s constitutional

architecture. This explains why, at the time of the Treaty of Rome (1957), an independent judicial body – the European Court of Justice – was endowed with enforcement powers for which there was no real precedent in the international sphere. As is known, the Court, acting at the request of the Commission, can rule that states have failed to comply with EU law. If states fail to abide by its judgements, the ECJ can impose financial sanctions, following an innovation introduced by the Maastricht Treaty. The system is not without flaws20 but overall it has enabled EU decisions to be taken seriously in national capitals. The Court’s mandate is particularly broad: it must “ensure that the law is obeyed”, and the Treaty provisions it must enforce often are fairly vague, to the point that it could be described as an “imperfect contract” in the economic theory sense, i.e. a contract in which a number of contingencies are missing because they cannot be anticipated.21

Deciding what constitutes a ‘measure having equivalent effect to a quantitative

restriction’ in the sense of Article 30 EEC or a prohibited state aid, for instance, involves a substantial margin of interpretation. Granted, the members of the Court are not free of any ties to their respective countries since they are appointed by national governments. However, various devices, such as the fact that decisions are taken by consensus or through a majority vote, and that the deliberations are shrouded in secrecy, make it relatively cost-free for

18 This story can be viewed as a confirmation of Rogoff’s model of delegation of monetary policy to a central banker that is more conservative than the government. See K. Roggoff, The optimal Degree Commitment to the Intermediate Target, Quaterly Journal of Economics, 1985, p. 1169-90. 19 ‘When exercising the powers and carrying out the tasks and duties conferred upon them by the Treaties and the Statute of the ESCB and of the ECB, neither the European Central Bank, nor a national central bank, nor any member of their decision-making bodies shall seek or take instructions from Union institutions, bodies, offices or agencies, from any government of a Member State or from any other body. The Union institutions, bodies, offices or agencies and the governments of the Member States undertake to respect this principle and not to seek to influence the members of the decision-making bodies of the European Central Bank or of the national central banks in the performance of their tasks.’ 20 G. Falkner, Is the EU a Non-Compliance Community? Towards ‘compliance for credibility’ and EU action for the protection of democracy in Europe, Les Cahiers européens de Sciences Po, 01, 2013. 21 R.Posner, Economic Analysis of Law, 3rd ed., Boston: Little & Brown, 1986.



14

European judges to deviate from their own country’s preferences22. Recurrent criticisms of the Court’s pro-integration bias23 can be read as confirming that it is not particularly constrained by its principals’ concerns. Describing it as a trustee of the common interest, rather than as a mere agent of the Member States, therefore seems appropriate.

Similar considerations also explain other “anomalies” in EU policy-making. Whereas

the rule is that regulatory principles are defined at the European level and implemented by domestic authorities, during the drafting of the EEC Treaty another process was preferred with regard to competition policy. In this area, a few basic principles - the prohibition of state aids, cartels and of the abuse by firms of their dominant position in their respective markets - were enshrined in the Treaty, while their implementation was entrusted to the newly created European Commission. This atypical choice can be explained by various kinds of factors. First, it was known that economic policy was strongly embedded in national cultures: countries like France or Italy had a strong tradition of state intervention in the economy, whereas the appeal of ‘Ordnungspolitik’ was strong in Germany. To make things worse, only Germany had an established antitrust policy at the time.24

Given these differences, there were good reasons to doubt that the Treaty’s vague

principles would be implemented in the same way throughout the Community and thereby ensure a level playing field.

In such a context, the idea of entrusting an independent body such as the European

Commission with discretionary powers made much sense. For this solution to be credible, however, the agent’s autonomy had to be protected. As is known, members of the Commission are supposed act in complete independence, while governments are explicitly prohibited from ‘[seeking] to influence them in the performance of their task’ (Article 245 TFEU). Commissioners owe their appointment and possible renewal of their mandate to their national government, and might thereby be incentivized to pay heed to their principal’s concern. A countervailing force, however, may be the collegiality principle: as decisions must in principle be taken (or at least be screened) by the Commission as a college, individual commissioners within the executive cannot be seen as taking their orders from their own country without risking their reputations. This helps to reduce the risk of capture by sectorial interests25 and, by the same token, makes the implementation of competition policy via Commission decisions a more credible option than the alternative of decentralized implementation.

22 R. Dehousse, The European Court of Justice: The Politics of Judicial Integration, Basingstoke: Macmillan, 1998. 23 See e.g. H. Rasmussen, On Law and policy in the European Court of Justice. A Comparative Study in Judicial Policy-Making, The Hague: Kluwer, 1985 or F.Scharpf, Legitimacy in the Multilevel European Polity, MPIfG Working Paper, 09(1), 2009. 24G. Majone, Regulating Europe, London: Routledge,1996, chapters 7 and 8. 25 It is fair to say that today the weakening of collegiality with the emergence of a presidential administration and the disappearance of voting in the Commision undermines the strength of the control mechanism. See H. Kassim et al., The European Commission of the XXIst Century, Oxford University Press, 2013. This has occasionally resulted in proposals to take competition policy out of the Commission’s control to entrust it to some independent anti-trust authority.



15

Another remarkable feature of EU policy-making is the fact that the Treaties have

delegated the power to initiate legislation almost entirely to the Commission. The latter’s proposals are moreover protected by the fact that unanimity in the Council is required to modify them (Article 293 (I) TFEU). This atypical mechanism has often been explained by the fact that the Commission’s composition and mandate make its long-term commitment to integration more credible than that of national governments. Concentrating agenda-setting powers in its hands is seen as a way to strengthen the drive towards integration generated by a wider use of majority voting26: had a similar right been granted to the Member States, they could have been tempted to renege on their commitment to European integration, thereby compromising the acquis communautaire27. However, this kind of ex post facto reasoning seems to place too much emphasis on potential divergences between the interests of the Commission and those of the Member States, and not enough on those that may exist between the interests of the Member States themselves. In fact, the Commission’s right of initiative and the ‘supranational veto’28 that protects it, originated in proposals tabled by Benelux countries at the Val Duchesse conference, where the Treaty of Rome was drafted. Fearing that they might be systematically outvoted due to the weight of big countries in the Council, they sought protection through the conferral of agenda-setting powers to a neutral umpire29. In other words, mistrust between state principals once more appears to have been a central factor in the politics of delegation.

As the above examples show, delegation of substantial powers to an independent

institution required acting as a trustee of the common interest, has frequently been used to overcome potential tensions in situations where national preferences and traditions substantially differed, and trust between the Member States was accordingly limited. Recently, however, another type of delegation has emerged, in which similar concerns have led to different outcomes.

4. Type II Delegation: EU Administrative agencies

The Growing Polycentricism of the EU Executive

One of the most interesting developments of the past two decades in European public

administration is the mushrooming of autonomous administrative structures of various kinds. While only two semi-autonomous agencies existed prior to the 1990s, there are 35 today. This figure includes the various organizations established under the label of ‘decentralized agencies’, as well as the ‘Union agencies’ set up in the second and third pillars, such as the

26 G. Tsebelis and A. Kreppel, The History of Conditional Agenda-Setting in European Institutions, European Journal of Political Research, 33, 1998, p. 41-71. 27 K. Lenaerts, Regulating the Regulatory Process: ‘Delegation of Powers’ in the European Community, European Law Review, 1993, p. 22-49; G. Majone op. cit., at 115. 28 J. Weiler, The Transformation of Europe, 100 Yale Law Journal, 1991, p. 2403-83. 29 H.J. Küsters, Les fondements de la Communauté économique européenne, Brussel : Labor/Luxembourg : EUROP,1990.



16

Institute for Security Studies, EUROPOL or EUROJUST.30 Although this decentralization is no longer new, it seems to have accelerated in recent years, with thirteen agencies created during the Prodi Commission’s mandate, followed by another increase in the wake of the financial crisis. In the 2000-2013 period, the number of administrative posts in the establishment plans of decentralized agencies increased fourfold, from 1486 to 6050, while the EU contribution their budget increased eightfold, from 95 million in 2000 to 775 million in the 2013 budget.31 The creation of European agencies was a fairly haphazard development, often prompted by external factors, such as food scares (for the Food Safety Authority), shipwrecks (for the Maritime Safety Agency) or the financial crisis (for the Banking Authority). None of the main institutional actors deliberately planned or defended these developments as the way forward. Proposals to establish an agency were often met with resistance from one side or the other and usually gave rise to fairly intense inter-institutional bargaining. Yet a phenomenon of such magnitude and regularity is likely to be a response to structural constraints.32

Thus far, the movement that led to the establishment of EU agencies has been analysed mainly along functional lines, using the principal-agent model.33 The need to bring expertise to the public policy process or to ensure its credibility feature prominently among the motivations attributed to those who promoted this new trend at the European level. However, despite the unquestionable relevance of this model, its standard form does not take into consideration some of the EU setting’s peculiarities, the most important one being the absence of a clearly defined “principal,” as discussed above. Failing to acknowledge the wide variety of actors that participate in the final decision would make it difficult to account for the many tensions surrounding agency creation, or the ambivalence of the Commission, which has long appeared reluctant to accept delegations of power and yet has continued to propose the establishment of new agencies. Similarly, the Commission’s insistence on the need to streamline the creation and functioning of agencies, resulting in a ‘common approach’ jointly defined in 2012 by the Parliament, the Council and the Commission,34 cannot be understood without a reference to the multiplicity of principals in the Union. Why would a single principal want to codify delegation practices?

Who are the Principals?

It is important to stress that the principals in question are not the same as in the ‘full

delegation’ model discussed above. The creation of delegated agencies is decided through legislative measures that require agreement between several institutions: first the

30 In contrast, this figure does not include the ‘executive agencies’ that were created in the wake of the Kinnock reform to relieve the Commission of routine executive tasks, and that enjoy far less autonomy. 31 European Commission, Communication to the European Parliament and the Council. Programming of Human and Financial Resources for Decentralized Agencies 2014-2020, COM(2013) 519 final of 10 July 2013. 32 See R.Dehousse, ‘Delegation of Powers in the European Union: The Need for a Multi-principals Model’, West European Politics, 31(4), 2008, p. 151-167 on which this section largely rests. 33 See e.g. G. Majone ‘The new European Agencies: Regulation by Information’, Journal of European Public Policy, 4(2) 1997, p. 262-75 or M. Thatcher and A. Stone Sweet. Theory and Practice of Delegation in Non-Majoritarian Institutions, West European Politics, 25(1), 2002, p. 1-22. 34 Joint Statement of 19 July 2012, available at http://europa.eu/agencies/documents/joint_statement_and_common_approach_2012_en.pdf.



17

Commission and the Council, and then the Parliament as well, following the expansion of the latter’s legislative prerogatives. Each can claim to play a meaningful role in the delegation process.

In various position papers, the Commission has implicitly, but nonetheless clearly

presented itself as the principal that must anticipate the possibility of delegating a share of its powers to autonomous bodies. It has repeatedly stressed the need to preserve ‘the unity and integrity of the executive function’, which it deems to be ‘vested in the chief of the Commission’.35 Similarly, in a 2005 draft Inter-institutional Agreement, agencies are described as mere auxiliaries that must ‘provide the Commission, in particular, with the experience and expertise it needs so that it can fully meet its responsibilities as the Community executive.’ The Commission’s legal service has forcefully defended this vision, which purportedly stems from the infamous Meroni ruling, whereby the European Court of Justice narrowly circumscribed the range of powers that may be delegated to other actors.36 The legal community tends to reason as if the creation of autonomous administrative agencies should be mainly considered a transfer of authority from the Commission to these bodies.37

This view, however, seems to rest on a fairly abstract view of the distribution of labour

between the various institutional actors that have a say in the delegation process; it pays insufficient attention to the actual transfers of power that have occurred whenever agencies were established. More often than not, the powers entrusted to European agencies were previously held by national authorities rather than by the Commission. Prior to the establishment of the European Medicines Agency (EMEA) or the Office for Harmonisation in the Internal Market (OHIM), marketing authorisation for medicinal products and trademark regsitration were handled by national bodies rather than the Commission. Thus, delegation entailed a vertical transfer of power (from the national to the EU level) rather than a horizontal one (from the EU institutions to specialised agencies). Outside of the ‘executive agencies’, instances of actual transfers of powers from the Commission to bureaucratic structures have remained very rare.38 Moreover, since the creation of agencies is decided through a legislative procedure, it requires agreement between actors of various types, each with their own interests. A proposal to this effect must be tabled by the Commission, accepted by national governments and, today, by the European Parliament. The Commission thus proposes the establishment of the agencies, greatly shapes their structure and powers,

35 European Commission, Communication on the operating framework for the European regulatory agencies, COM(2002)718 final of 11 December 2002. 36 Meroni v. High Authority, [1957] ECR 133. 37 See e.g. E. Chiti, The Emergence of a Community Administration: The Case of European Agencies, Common Market Law Review, 37, 2000, p. 309-43; D. Géradin, R. Munoz and N. Petit (eds). Regulation through Agencies in the EU. A New Paradigm for European Governance, Cheltenham: Edward Elgar, 2005 ; K.Lenaerts, op. cit.; X. Yataganas, Delegation of Regulatory Authority in the European Union: the Relevance of the American Model of Independent Agencies, Jean Monnet working Paper 3(01), 2001, and R. Dehousse Misfits: EU Law and the Transformation of European Governance, in C. Joerges and R. Dehousse, Good Governance in Europe’s Integrated Market, Oxford university press, 2002, p. 207-29 for a critique. 38 The above-mentioned European Agency for Reconstruction was exceptional in that respect.



18

and often provides them with their first director. As it has gained co-decision powers, the European Parliament has also become a major participant in the politics of agency design.39

The Contribution of the Multiple-Principals Approach

Recognising the plurality of principals enables us to understand some key structural features of the decentralized bodies that have been established. Tensions among principals explain a number of stable elements in the otherwise haphazard process that has led to the mushrooming of various kinds of agencies. These tensions have affected both delegation decisions and the institutional design of the agencies.

First, the multiplicity of principals has largely shaped the politics of agency creation.

Each of the principals is somewhat reluctant to relinquish power (as with all principals, in general). However, their main concern is not what the P-A literature describes as ‘agency drift’, that is, the idea that agents can pursue a political agenda that differs from that of their principal. What the principals fear most is the emergence of a variant of ‘political drift’ in which agencies are somehow ‘captured’ by one of their institutional rivals in the leadership contest. Thus, as a rule, even when national governments accept the need for enhancing cooperation at the European level, they tend to oppose granting significant powers to the Commission. Similarly, when the first European agencies were established the national governments insisted that they be subject to intergovernmental control so as not to threaten national administrations. EU agencies were therefore established as the hubs of networks bringing together national administrations.40 For its part, the Commission is naturally inclined to maximise its own power like most bureaucratic structures,41 and fears the emergence of potential rivals that will be more exposed to member states’ influence. From its perspective, delegation is often only a second-best alternative, which it will accept only if convinced that an extension of its own powers is unlikely to be approved by the Council. Thus, in the field of food safety, it supported the establishment of an independent authority only once it realized that the national governments were unwilling to enlarge its prerogatives.42 The Commission’s 2001 White Paper on European Governance clearly illustrates this ‘official’ point of view. In typical fashion, it highlights the advantages of agencies (sectoral expertise, increased visibility and cost savings), but dwells at greater length on the limits that should surround the delegation of powers if the balance between the institutions is to be respected.43 This

39 D. Kelemen, The Politics of ‘Eurocratic’ Structure and the New European Agencies, West European Politics, 25(4), 2002, p. 93-118. 40 R. Dehousse Regulation by Networks in the European Community: The Role of European Agencies, Journal of European Public Policy, 4(2), 1997, p. 246-61. 41 W. A. Niskanen, Bureaucracy and Representative Government, Chicago: Aldine Atherton, 1971. 42 D. Kelemen op. cit. 43 ‘Agencies can be granted the power to take individual decisions…but cannot adopt general regulatory measures’, nor can they ‘be given responsibilities for which the Treaty has conferred a direct power of decision on the Commission (for example, in the area of competition policy).’ They ‘cannot be granted decision-making powers in areas in which they would have to arbitrate between conflicting public interests, exercise political discretion or carry out complex economic assessments.’ Finally, of course, ‘agencies must be subject to an effective system of supervision and control’ (European Commission, European Governance. A White Paper, COM(2001) 428 final, 25 July 2001, at 24.



19

lukewarm view is often shared by the European Parliament, which fears being deprived of its hard-won legislative powers by the regulatory agencies, despite the fact that its own influence over the Commission has been growing over the past decade. It therefore generally favours a model in which ‘the autonomy of the new regulatory agencies …[is] exercised under the direct supervision of the Commission and monitored politically by the European Parliament’.44 From this perspective, the ‘enemy’ is represented by the member states’ intergovernmental bias since, unlike the Commission, national representatives are not under the Parliament’s authority.

In other words, one finds at this level a pattern similar to the one that characterizes the

‘full delegation’ model, in which the politics of delegation are largely shaped by mistrust between the principals.² This explains why agency creation has often led to protracted negotiations and elaborate compromises between the various branches of the legislature, and why the institutional architecture of EU regulatory agencies is complex. Although the establishment of these new structures has often been criticized as a somewhat slapdash process45, there are a number of recurrent features in their configuration and functioning, all of which appear to be directly linked to a bid to prevent the capture of the newly created structure by one of the other principals.

In the principal-agent literature, for instance, the power to appoint the heads of an

agency’s executive is regarded as a standard control mechanism. At the EU level, this power is generally fragmented. Most often, directors are appointed by the agency’s administrative board based on a proposal from the Commission; in other cases, he or she is appointed by the Commission based on a proposal from the administrative board or by the Council on the basis of a list of candidates drawn up by the administrative board or the Commission.46 The Parliament’s role in the appointment process is less prominent, except for a few agencies whose work is of direct interest to the general population, such as the European Food Safety Authority (EFSA) or the European Centre for Prevention and Control of Diseases (ECDC). The essence of the compromise is therefore clear: to avoid concentrating the power to appoint in any one of the would-be ‘principals.’ Likewise, while there are several possible variants as regards the composition of agencies’ administrative boards, in all but three cases they are comprised of at least one representative from each member state. This occasionally leads to rather peculiar situations. For example, Hungary and Malta are entitled to appoint one person to the board of the European Maritime Safety Agency or the European Railway Agency, respectively, even if their representative’s expertise in those fields can be assumed to be limited.47 In the three agencies in which not all states have representatives – the Food Safety Authority (EFSA), the Institute for Gender Equality (EIGE) and the Agency for the

44 European Parliament, Report on the Communication from the Commission: ‘The Operating Framework for the European regulatory agencies’ (COM(2002) 718 – 2003/2089 (INI), drafted by Ms. Teresa Almeida Garrett, EP Doc. 15-1471/2003 final, 4 December 2003. 45 See e.g. European Commission, Communication on the operating framework for the European regulatory agencies, COM(2002)718 final of 11 December 2002. 46 The Council enjoys more autonomy for so-called “Union agencies”, created in the fields of foreign policy or justice and home affairs, where it may be regarded as the main principal. 47 The only railway line in Malta was closed in 1931!



20

Cooperation of Energy Regulators (ACER) – the Management Board is relatively smaller and includes members who are supposed to act in the public interest, or represent Member States on a rotating basis. In these three cases, however, the founding regulations foresee the creation of advisory bodies, which are bigger in size and in which all Member States are represented.

In similar fashion, principals seem to have found it appropriate to confine agencies to a

secondary role, at least formally. The majority of the organizations created since 1990, such as the European Environment Agency in Copenhagen, have an information-gathering mission and may assist the Commission in implementing programmes and policies, like the European Training Foundation in Turin. They may even go so far as to prepare decisions to be taken by the Commission, as is the case for the London-based European Medicines Agency (EMEA). Yet most have been denied any independent decision-making power. As a result, only a handful of agencies, such as the Office for Harmonisation in the Internal Market in Alicante, the Community Plant Variety Office in Angers or, more recently, the Aviation Safety Agency (EASA) in Cologne, have been granted the power to make individual decisions on the grounds that such decisions do not entail significant discretionary powers. Needless to say, the power to adopt formal rules of any kind is always denied to those bodies. The EASA, however, has been entitled to adopt non-binding ‘guidance material’48 which, coupled with the authority to issue individual decisions, approaches independent regulatory power, as demonstrated by the Commission’s experience in the field of competition policy. Moreover, when it advises the Commission on the rules to be adopted to implement air safety regulations, the technical part of its opinions – in particular, the part covering the construction, design and operation of aircraft – cannot be altered by the Commission without prior collaboration with the Agency (Article 12b). The fact that the Agency’s influence has been somewhat disguised, however, strongly suggests there is a deep reluctance to formally grant regulatory authority.

Be that as it may, the absence of formal authority does not necessarily mean that

agencies are deprived of any influence whatsoever. For instance, the Commission systemtically rubber-stamps EMEA recommendations.49 This is hardly surprising. If an institution pooling the best expertise available at the European level warns against the dangers of a given medicine, the ‘political power’ cannot ignore its advice without taking substantial risks. More generally, if regulatory bodies manage to acquire (and preserve) credibility, they may acquire significant influence even when all they have is mere advisory or information functions.50 But are all European agencies, with their multiple principals, strong enough to acquire that type of influence?

Finally, the number and variety of control devices used to monitor EU agencies’ action

is impressive. Agencies are required to present an annual activity report describing how they

48 Art 13 (b) of Regulation 1592/2002, OJ L 240 of 7 September 2002. 49 B. Hauray, L’Europe du médicament, Paris : Presses de Sciences Po, 2006. 50 G. Majone The new European Agencies: Regulation by Information, Journal of European Public Policy, 4(2), 1997, p. 262-75.



21

implemented their annual programme. EU funding is included in the Union’s general budget, which the Commission prepares and the Council and Parliament approves. For some time now, the European Parliament has used its ‘power of the purse’ to make its voice heard and to influence agency activies.51 Furthermore, agencies are subject to the authority of the Commission’s Financial Controller, whereas the European Parliament takes a final decision on whether to discharge their budgets upon the Council’s recommendation.52 Additionally, judicial oversight mechanisms have been implemented for agencies with the power to adopt individual decisions, such as the OHIM or the EASA. The Ombudsman has also used its oversight authority to recommend that agencies adopt rules on transparency and access to documents.

In the principal-agent literature, accountability mechanisms are mostly viewed as a

method by which the principal can ensure that the agent follows the instructions it has been given rather than developing its own priorities. As suggested above, in a system with multiple principals one of their main fears is that an agency will be subjected to undue influence by a rival in the contest for political leadership. The number and variety of accountability mechanisms makes this kind of ‘capture’ unlikely. The fact that the Commission and the national administrations contribute to their internal functioning (rather than act merely as forces of external control) makes it more difficult for agencies to adopt a policy that would cross any red lines defined by their principals. Indeed, in her study of management boards, Madalina Busuioc found them to be characterized among other things by ‘a lack of interest in overall agency performance; a bias towards a national outlook not aligned with their European role; the presence of conflicts of interests; a focus on administrative detail.’53 This of course makes it fairly unlikely that strong regulatory agencies will emerge.

In sum, the tug of war between the three big EU institutions has resulted in the

emergence of a fairly weak delegation model, which we will label as ‘type 2 delegation’ for lack of a better name. None of the agencies created during that period can be depicted as a powerful regulator. Even though some have acquired a degree of influence in their field, their authority is regularly challenged by some of their interlocutors, and multiple accountability channels have been established in order to ensure that they do not overstep the boundaries assigned to them. It is true that this type of polycentricity is not unheard of at the domestic level, where tensions may exist between, for example, the parliament and the executive when the creation of autonomous agencies is debated.54 Yet there is, at the very least, a difference in degree. In a system of multi-level governance such as the EU, polycentricity may be regarded as the cornerstone of the structure, and often prevents the emergence of strong agents.

51 Brinkhorst L.J., The Future of European Agencies: A Budgetary Perspective from the European Parliament, in Alexander Kreher (ed.), The New European Agencies, EUI Working Paper RSC n. 96/49, 1996, p. 75-81. 52 Financial Regulation 1605/2002 of 25 June 202, OJ L 248 of 16 September 2002. 53 M. Busuioc, European Agencies and their Boards: Promises and Pitfalls of Accountability beyond Design, Journal of European Public Policy, 19(5), p. 719-36. 54 see e.g. M. Thatcher and A. Stone Sweet op. cit. and M. Shapiro, The Problem of Independent Agencies in the United States and the European Union, Journal of European public policy, 4(2), 1997, p. 276-91.



22

5. A Model in Crisis?

Growing Signs of Resistance

Since the beginning of the 1990s, the willingness of national governments to accept further transfers of powers to the European level has considerably eroded. Signs of their growing impatience with what they see as an unlimited increase in the powers of the Commission are plentiful, and have led to a regular search for counterweights of various kinds to its power.

The ‘pillar structure’ of the Maastricht Treaty was undoubtedly the first expression of

this tendency. While member states accepted the necessity of common action in areas that are traditionally the preserve of the state, such as foreign policy, security and justice, they refused to allow EU supranational institutions to adopt a role commensurate with the one they play in the first (economic) pillar. In these areas, the only forms of states envision fall within the more traditional intergovernmental framework, with the European Council providing leadership. Although the Treaty created the European Central Bank, whose independence is well protected, as we saw, several features of its economic policy chapter illustrate states’ reluctance to delegate new powers to supranational bodies. In the excessive deficit procedure, for instance, the Commission has seen its 'conditional agenda-setting' role curtailed since it can only issue recommendations which, unlike its proposals, the Council can alter with a qualified majority. The importance of this nuance was evident in 2003, when France and Germany, which were running higher deficits than permitted by the Treaty, found sufficient support among their peers to avoid the Commission’s suggested censure. Additionally, the ability to bring infringements of the treaty before the Court of Justice was excluded in this area. National governments’' aversion to centralization was also illustrated by their refusal in 2005 to authorize an increase in the oversight powers of Eurostat, the EU statistical office that is technically a directorate-general of the Commission. It is of course impossible to tell whether this change would have contributed to an earlier uncovering of the Greek public accounts forgery which the Papandreou government revealed in 2009. What is clear, however, is that member states’ unwillingness to accept any intrusive European control in the field of economic policy has created conditions that allowed a series of drifts to go either unnoticed or unsanctioned, thereby facilitating the eruption of the sovereign debt crisis.

Typically, ad hoc structures have been set up when the need for steadier steering has

arisen. For example, when European foreign policy was seen to be suffering from a severe lack of analytical and planning structures capable of inspiring a common vision of international issues, defining potential joint action and guiding its implementation, a special policy unit was established and placed under the authority of an autonomous individual with limited powers – the High Representative for the Common Foreign and Security Policy. Member states only reluctantly accepted that this office had to somehow be brought closer to the Commission for the EU to get its act together. A similar scenario has unfolded in economic policy matters. Here again, the compromises of Maastricht proved unstable. To



23

avoid any threat to the independence of the European Central Bank, a mere informal discussion forum (the ‘Eurogroup’) was set up for the finance ministers of the countries participating in the single currency, and was later strengthened by the creation of a more stable presidency. The desire to define a viable intergovernmental alternative to the transfer of powers to the Commission – characteristic of the Community method – could not be clearer. Similarly, one of the Lisbon Treaty’s main innovations, which was the creation of an office of president of the European Council, can be seen as a deliberate attempt to create some sort of intergovernmental counterpart to the President of the Commission. For all his soft-spoken style, the first holder of that office, Herman Van Rompuy, has clearly stepped into the initiative and mediation role traditionally devolved to the Commission.55

A similar phenomenon can be observed with regard to policy instruments. The wave of

harmonisation that marked the completion of the internal market has been succeeded by a new phase characterised by working methods that impose fewer constraints on national administrations: benchmarking, peer review and mutual monitoring are key techniques in these ‘new modes of governance’, which make extensive use of networks of various types and profess greater openness to civil society in public policy-making.56 Taken together, these moves reflect a clear desire to break with the broad delegation of powers that is distinctive of the Community method, and a clear propensity to favour what Helen Wallace has defined as ‘intensive transgovernmentalism’.57

While an examination of the causes behind these changes is beyond the scope of this

contribution, one can note that they coincide with changes in popular perceptions of the European Union. Opinion polls have unanimously confirmed that the ‘permissive consensus’ that enabled the European venture to launch58 is now nothing but a memory, even though there is debate over the dominant feature in citizens’ attitude towards the EU: indifference or growing dissent.59 At first sight, the succession of financial crises Europe has experienced since 2008 seems to have compounded this centrifugal tendency. Under the pressure of financial markets, the European Council has seen its role considerably enhanced. The number of meetings has risen dramatically (9 meetings were held in 2011, i.e. about one every six weeks) while matters normally addressed at Council of Ministers level were

55 P. de Schoutheete, The European Council and the Community Method, Notre Europe Policy Paper, N. 56, 2012, available at http://www.notre-europe.eu/media/europeancouncil_ph.deschoutheete_ne_july2012_01.pdf?pdf=ok 56 L. Boussaguet and S. Jacquot, Les nouveaux modes de gouvernance, in R. Dehousse (ed.), Politiques européennes, Paris : Presses de Sciences Po, 2009, p. 409-28. 57 H. Wallace, The Institutional Setting in H. and W. Wallace (eds.), Policy-Making in the European Union, Oxford University Press, 2000, p. 3-37. 58 L. Linberg and S. Scheingold, Europe’s Would Be Polity: Patterns of Change in the European Community, Englewood Cliffs, NJ: Prentice Hall, 1970. 59 L. Hooghe and G. Marks, A Postfunctionalist Theory of European Integration: From Permissive Consensus to Constraining Dissensus, British Journal of Political Science, 39(01), January 2009, p 1-23; S. Duchesne, E. Frazer, F. Haegel and V. Van Ingelgom, (eds), Citizens' Reactions to European Integration Compared. Overlooking Europe, Basingstoke: Palgrave Macmillan, 2013.



24

increasingly referred to the heads of state and government. According to the prevailing view, the Eurozone crisis has therefore reinforced the shift towards intergovernmentalism that characterizes the post-Maastricht period. The reality, however, seems to be somewhat more complex, as we shall now see.

The Eurozone Crisis and the Politics of Delegation

Against all odds, the reform process initiated in November 2011 with the adoption of

the ‘Six Pack’ (the Eurospeak name for a package of legislative decisions adopted in the wake of the crisis), as well as the 'Fiscal Compact' imposed by Chancellor Merkel in the ensuing months and the ‘Two Pack’, in force since May 2013, has significantly strengthened the Commission's hand in the coordination of macroeconomic policy. The scope of its oversight powers has been extended to include a larger series of macroeconomic indicators than member states' budget deficits, as the Irish and Spanish crises have shown that private indebtedness, for instance, could also be a source of major difficulties. Monthly monitoring is foreseen for countries against which an excessive procedure has been opened. The authority of Commission recommendations on Eurozone countries has also been strengthened since the Council is assumed to support them unless a qualified majority is found to overturn them (‘reverse qualified majority’). And if it were not for the British Prime Minister's refusal to endorse a modification of the EU Treaty, thereby limiting the possibilities of reform, the German government and its allies would in all likelihood have secured an extension of the Court's oversight powers in the realm of economic policy.

At the time these measures were taken, it was not clearly understood that they

significantly improved the Commission's position. All eyes were turned to the European Council, in which the 'Merkozy' tandem appeared to be decisive. It may seem that the Commission's influence in the negotiations was limited. However, this overlooks the fact that some of the innovations, such as the 'European semester', which foresee an ex ante control of national budgets by the Commission to ensure their compatibility with national governments’, have for many years been advocated by its ECFIN directorate-general.60 The sovereign debt crisis simply made this kind of preventative monitoring much more appealing to countries that were anxious to see their partners' fiscal discipline monitored more closely. Since the enactment of the new provisions, a number of countries have experienced Commission criticism. Their reactions suggest that they were somehow caught off-guard by the intrusive character of Commission oversight. In May 2013, French President François Hollande vehemently reacted to a Commission hint that a more radical pension reform would be needed: France, he said, is a sovereign country, and would make its own decisions about how to reach its public finance targets. Later that year, the Italian government described as inappropriate the scepticism expressed by Commissioner Olli Rehn regarding Italy's ability to reduce public spending.61 Attacks on the unelected character of the Commission often

60 P. Copeland and S. James, Policy windows, ambiguity and Commission entrepreneurship: explaining the relaunch of the European Union's economic reform agenda, Journal of European Public Policy, 21(1), 2014, 1-19. 61 ‘Letta: Rehn non può permettersi scetticismo sull'Italia’, La Repubblica, 3 December 2013.



25

featured prominently in the reactions. 'Who knows Mr. Rehn?' asked Belgian minister Paul Magnette in January 2012,62 after the Commission expressed doubts about the Belgian government’s debt reduction programme.

There is of course a mild irony in this kind of rhetoric. While they clearly intend to

underscore the weak legitimacy of the European executive, such remarks actually point to the main reason underpinning the delegation of oversight powers to supranational bodies. If the Commission was electorally accountable, there is little chance it would qualify as a credible monitor of national policies. Be that as it may, the gradual emergence of a stronger role for the Commission in economic policy shows that we should avoid rushing to conclusions in our interpretation of the institutional consequences of the crisis. There is no dispute that the European Council made all of the major decisions, as would arguably be expected in a period of crisis. But it does not follow that in the longer run it will carry on playing such a prominent role. Interestingly, in an interview given in September 2013, European Council President Herman Van Rompuy, who has been one of the chief architects of the EU's response to the crisis, similarly suggested that ‘[o]ne of the paradoxes of the crisis is that the intergovernmental dimension embodied by the European Council has contributed to the strengthening of the role of the Community institutions.’63

'Banking union', i.e. the birth of an European banking regulation policy, provides

another interesting example of the politics of delegation during a period of crisis. As is known, the European banking sector was severely hit by a series of crises from 2008 onward: the credit crunch following the bankruptcy of Lehmann Brothers, the turmoil deriving from banks' exposure to the sovereign debt of countries like Greece, and the real estate bubbles in Ireland and Spain all contributed to undermining public confidence in European banks. As a result, the idea gained ground that in a system where capital can move freely, a degree of centralized control over the banking sector is required.

The first step in this direction, the establishment of an European Banking Authority in January 2011, is in many respects a perfect example of a Type II delegation. Its main decision-making body is the board of supervisors, composed of the EBA's Chairperson and of representatives of the 28 national supervisory authorities, even though a management board of six members has been set up. As noted by a careful observer, its governance structure `is highly problematic, as it gives each member state equal weight in crucial decision-making and does not give voting power to any representation of the EU interest beyond individual member states.'64 Its powers are limited. True, it is tasked with the elaboration of binding technical standards in order to create a single set of harmonised prudential rules for financial institutions throughout the EU. But its regulatory powers have been tailored according to a strict reading of the Meroni doctrine.65 Several parts of its 62 ‘Mais qui est Olli Rehn ?’, Le Monde, 27 February 2012. 63 ‘L’idée européenne, une convergence d’intérêts, un choix de valeurs’, La revue des deux mondes, September 2013, p. 15-31 at 17 (author’s translation). 64 N. Veron ‘London and the new European Supervisory Authorities’, Financial World, 7 April 2011. 65 M. Busuioc, ‘Rulemaking by the European Financial Supervisory Authorities: Walking a Tight Rope’, European Law Journal, 19(1), 2013, p. 111-25.



26

founding legislation stress that it acts by delegation only; its standardization powers can even be revoked without amending the regulation.66 The Authority must also promote convergence of supervisory practices to ensure a harmonised application of prudential rules. In other words, one of its main tasks is to regulate the national supervision authorities, which remain the primary regulators. It can make individual decisions addressed to individual financial institutions in exceptional circumstances only. Finally, the EBA has been given a mandate to assess risks and vulnerabilities in the EU banking sector through, in particular, pan-European stress tests. The first tests it conducted in July 2011 were severely criticized. It was alleged that the EBA contemplated unduly optimistic hypotheses for political reasons; as a result, only a handful of banks, which were all located in small countries, failed the exercise. Not only were the tests of limited value, but they also exposed the weakness of the Authority. Indeed, the Spanish press largely blamed the poor performance of local banks on the fact that, unlike other large countries, Spain had no representative on the EBA’s management board.67

Whether the criticism was justified is irrelevant. It makes clear that because of its weak powers and of the risk of capture to which it was exposed, the credibility of the European Banking Authority was not sufficient to allay market fears during a severe crisis. Thus in the Summer of 2012, when faced with a further deterioration of credit conditions, Member States decided that it was ‘imperative to break the vicious circle between banks and sovereigns’.68 In exchange for agreeing to share responsibility for rescuing European banks, if need be, Germany insisted that the supervision of European banks be taken out of the hands of national authorities to be entrusted to a strong European regulator. Member states therefore decided that the European Central Bank should become the anchor of the single supervisory mechanism on the basis of article 127(6) of the Lisbon Treaty. Two considerations seem to have contributed to this choice. First, the ECB had since the beginning of the crisis demonstrated an ability to act swiftly and effectively – which it would confirm in September of the same year with the adoption of OMT scheme.69 Second, the interdependence created by the single currency provided a strong incentive to act at the Eurozone level, further reinforced by the uncooperative behaviour displayed by British Prime minister David Cameron a few months before, when the ‘fiscal compact’ was being negotiated.

This is not the place to provide a detailed analysis of how the ECB will fulfil this new task, which is replete with difficulties. But this development, together with the beefing up of the Commission’s surveillance powers in economic policy, provides a striking illustration of the politics of delegation in a period of crisis. In both cases, despite a political climate adverse to the delegation of powers, supranational institutions have seen their powers greatly enhanced in areas of strategic importance. How can one explain this sudden shift? The most reasonable hypothesis lies in the very gravity of the crisis and in the deep mistrust

66 See Art. 12 of Regulation (EU) No 1093/2010 of the European Parliament and of the Council of 24 November 2010 establishing a European Supervisory Authority (European Banking Authority), Official Journal L 331, 15/12/2010 67 ‘La Tercera bofetada de la autoridad bancaria europea’, El Pais, 30 October 2011. 68 Euro Area Summit Statement of 29 June 2012. 69 On which see above section 2.



27

that existed among European governments. Fiscally orthodox members of the Eurozone were aware of being intimately tied to the countries with troubled public finances. Having accepted a series of financial aid packages of unprecedented magnitude, they demanded in exchange a series of tight control mechanisms aimed at ensuring that all commitments would be honoured. The Commission appeared as the natural seat for an oversight function which is in many respects similar to its role as guardian of the treaties. Direct enforcement by other countries was unthinkable, as witnessed by the growth of anti-German feelings in several Southern European countries. Similarly, in the banking field, the establishment of a strong European regulator was a precondition to the creation of some mutualisation of the guarantees offered to investors. In both cases therefore, a ‘full delegation’ model, i.e. the transfer of important regulatory powers to an independent authority whose commitment to the policy objective appeared more credible, emerged as the most adequate solution.

Conclusion

This article has attempted to show the relevance of the principal-agent model to understanding some key features of the EU institutional system. We have seen that the explanatory power of the model is quite strong but that it needs to be supplemented by considerations linked to the multi-level character of the EU political system. First, the European Union is a polycentric system, deprived of a strong center of power. The rivalry that exists between states, and its translation into a competition between the main political institutions (Commission, Council and Parliament), shapes most of delegation politics. Secondly, mistrust between distinct principals may result in two contrasted patterns of delegation: a full transfer of powers to independent agents (of which the European Central Bank is currently the best example) or the creation of weak agencies that can be kept under tight control.

This multi-principals approach enables us to understand one of the paradoxes of

recent times: why in a period of crisis when centrifugal forces appear to be stronger than ever, European governments have nonetheless accepted further transfers of authority to independent supranational bodies. The explanation appears to have little to do with federalist ideology. The emergence of a strong European umpire may be nobody’s first choice, but given the mistrust that has developed among European states, it is deemed preferable to its alternatives. If that is the case, then in all likelihood, far from being a thing of the past, ‘full delegation’ still has a bright future.

Documents

Accueil | Sciences Po Centre d'études européennes